PROJECT SUMMARY/ABSTRACT Adverse drug reactions are a leading cause of death, and thousands of additional patients are prescribed medications which provide no benefit. Pharmacogenomics has allowed the discovery of genetic variants impacting response or toxicity for hundreds of drugs, but such information has infrequently been clinically utilized. Implementation has been hampered by poor physician knowledge, limited avenues for testing, delays in receipt of results, and informatics barriers to genomic implementation. There is also skepticism regarding the clinical utility of pharmacogenomics, underpinning the need for randomized examination of clinical efficacy. This was the genesis for our interest in proposing this project, which aims to understand whether a novel decision-support tool delivering patient-specific pharmacogenomic information can demonstrate that genetically-informed prescribing reduces inappropriate medication use in the perioperative and critical care settings. We previously developed and studied the tool, called the Genomic Prescribing System (GPS), among outpatient physician-patient pairs at our institution over the past five (5) years. GPS incorporates preemptively- obtained patient-specific pharmacogenomic results and translates these into clinical decision support summaries. The significance of this model is that it addresses several of the primary barriers to genomic implementation/dissemination?the need for instantaneous access to results at the point-of-care, translation of genomic information into decision-making logic, and provider education/decision support about genomics. We are interested to know whether our conceptual framework for adoption and use of pharmacogenomic information has unique mediators or unanticipated barriers in the perioperative setting, where anesthesiologists and critical care providers make many high-stakes prescribing decisions very rapidly. Additionally, in the perioperative setting the idea of pre-identifying patients who have genomic predisposition to increased medication risk has the potential for added clinical value that is greater than that for screening in other clinical populations, because for many perisurgical patients, it will be the first time having an operation (and thus the first time being exposed to various associated perioperative drugs). Our hypothesis is that a medical implementation model for personalized care that makes relevant pharmacogenomic information instantaneously accessible at the time of prescribing will reduce the use of inappropriate and high risk medications in patients for whom pharmacogenomic results are known. This hypothesis is based on the premise that the efficacious clinical translation of genomic discovery will be mediated by both systems/technology changes and changes in clinician behaviors. Impact on public health will result not only from the technology we employ but from our understanding of decision-making processes involved in promoting and adopting risk-reductive behavior in the era of precision medicine.